Are those the Hermite polynomials?
Ah. In that case, I'd go for the Rodriguez representation. Are you doing the probability version, or the physics version of the Hermites?
Look at this thread. It works out if you take the derivative correctly, which I'm not sure you did.